Risk for substance use (SU) including alcohol and other drug use, builds cumulatively over the course of adolescence culminating at a point of peak risk in late adolescence. Understanding the emergence of SU and related problems in late adolescence requires a characterization of the early adolescent risk context. SU information processing is a proximal cause of SU that mediates a broad range of more distal risk and protective factors. Current cognitive models of addiction distinguish controlled (slow deliberative and reflective) and automatic (fast and impulsive) processing and appraisal of SU information and posit that these processes jointly influence SU. Further, the role of information processing in SU is moderated by general capacity for self- regulation (i.e., executive functioning), which confers added risk. The literature supports these models for both adolescents and adults, yet very little of this research has taken a developmental perspective. Consequently, how automatic and controlled SU information processing develops, and whether these processes operate differently with age and stage of use is poorly understood. This is a critical gap in the literatur, as well- validated developmentally informed etiological models are crucial for the advancement of effective preventive interventions. Social context (e.g. peers, parents) and personality are strong determinants of SU, in part because of the role they play in shaping SU information processing. Social context changes dramatically from early to late adolescence, yet there has been little consideration of how these shifts may affect the development of SU information processing across stages of adolescence and stages of use. Indeed, no long- term longitudinal studies have followed youth from SU initiation, escalation, and emergence of problems to address these issues. Furthermore, neuroadaptations to SU change the motivational significance of SU, and likely impact reciprocal associations between SU and SU information processing, yet no research has examined such bidirectionality across adolescence. The proposed project addresses these gaps. This is a continuation of a study of alcohol and drug use in a community sample first assessed at ages 10-13 years. The proposed project will follow the sample into late adolescence (mean ages 18, 19, and 20 years) for 3 additional multi-method assessments that will include laboratory tasks and questionnaires. A critical feature of our longitudinal design is the ability to model multiple etiological influences, and pathways (mediation, moderation, reciprocal associations) and to compare directly how each may shift across age and stage of use (initiation, heavy use, problem use). The project will examine: 1) The changing role of automatic and controlled SU information processing; 2) Whether these processes mediate the influence of social context (parental SU- specific socialization, peer norms) and personality (reinforcement sensitivity) on SU; 3) Reciprocal associations between SU information processing and SU.